1. Field of the Invention
The present disclosure relates to methods and compositions for inhibiting angiogenesis.
2. Description of Related Art
β2-Glycoprotein 1 (β2GP1), also known as apolipoprotein H, is a 50 kDa plasma protein that is an abundant plasma glycoprotein found both free and associated with lipoprotein (Polz et al., FEBS Letters 102:183-186, 1979; Wurm, H., Int. J. Biochem. 16:511-515, 1984). Although the precise physiological role of β2GP1 is not known, in vitro studies suggest that it likely functions as a natural anticoagulant. β2GP1 has been shown to inhibit intrinsic pathway activation (Schousboe, I., Blood 66:1086-109, 1985; Schousboe and Rasmussen, Throm. Haemostasis 73:798-804, 1995), tenase (Shi et al., Throm. Haemostasis 70:342-345, 1993), and prothrombinase activities (Nimpf et al., Biochim. Biophys. Acta 884:142-149, 1986; Goldsmith et al., Brit. J. Heamatol. 87:548-554, 1994) on the surface of activated platelets and synthetic phospholipid vesicles. β2GP1 also inhibits the activity of activated protein C on procoagulant surfaces (Matsuda et al., American Journal of Hematology 49:89-91, 1995; Mori et al., Throm. Haemostasis 75:49-55, 1996). Plasma levels of β2GP1 have also been shown to fall during disseminated intravascular coagulation (Brighton et al., Brit. J. Heamatol. 93:185-194, 1996), which is consistent with the consumption of β2GP1 in this thrombotic process.
β2GP1 inhibits ADP-induced platelet aggregation (Schousboe, I., Throm. Res. 19:225-237, 1980; Nimpf et al., Atherosclerosis 63:109-114, 1987) and participates in the etiology of several thrombolytic diseases (see, e.g., Brighton et al., Brit. J. Heamatol. 93:185-194, 1996; Asherson and Cervera, J Invest Dermatol 100(1):21S-27S, 1993; Mackworth-Young, C., Immunol. Today 11:60-65, 1990; Kandiah et al., Lupus 5:381-385, 1996). β2GP1 preferentially binds to surfaces bearing negatively charged phospholipids. β2GP1 binds the negatively charged lipids phosphatidylserine (PS) and cardiolipin (CL), and regulates thrombosis by its ability to compete for the assembly of coagulation factors on PS-expressing platelet and endothelial cell membranes.
β2GP1 binds endothelial cells with high affinity, particularly endothelial cells that express acidic phospholipids (Ma et al., J. Biol. Chem. 275:15541-15548, 2000; Del Papa et al., J. Immunol. 160:5572-5578, 1998; Del Papa et al., Arthritis Rhuem. 40:551-561, 1997; Del Papa et al., Clinical & Experimental Rheumatology 13:179-185, 1995), and undergoes specific proteolytic cleavage (Horbach et al., Throm. Haemostasis 81:87-95, 1999). Other important in vivo targets of β2GP1 interaction are apoptotic or necrotic cells, where anionic phospholipids normally located in the inner side of the cell membrane become surface-exposed and serve as targets for β2GP1 binding (Balasubramanian and Schroit, J Bio. Chem. 273:29272-29277, 1998; Balasubramanian et al., J. Biol. Chem. 272:31113-31117, 1997). Subsequent to binding, the protein undergoes a conformational change (Wagenknecht and McIntyre, Throm. Haemostasis 69:361-365, 1993; Borchman et al., Clin. Exp. Immunol. 102:373-378, 1995; Lee et al., Biochim. Biophys. Acta 1509:475-484, 2000) that is recognized by a specific lipid/β2GP1 dependent receptor on the surface of phagocytes (Balasubramanian and Schroit, J. Biol. Chem. 273:29272-29277, 1998). Similarly, the association of β2GP1 with lipids can serve as antigens for antiphospholipid/β2GP1 autoantibodies (aPLAs) that are associated with systemic lupus erythematosus and antiphospholipid syndrome (McNeil et al., Proc. Natl. Acad. Sci. (USA) 87:4120-4124, 1990; Bevers et al., Throm. Haemostasis 66:629-632, 1991; Galli et al., Lancet 335:1544-1547, 1990). Although the structural rearrangements that are important for lipid-dependent phagocyte recognition and the generation of antiphospholipid antibodies are unknown, it is unequivocal that both the protein and the target membranes undergo critical changes in their conformation.
β2GP1 is a single-chain glycoprotein composed of 326 amino acid residues and consists of four complement control protein (CCP) modules (domains I through IV), as well as a distinct C-terminal domain V (Steinkasserer et al., Biochem. J. 277:387-391, 1991). From the crystal structure of β2GP1 (Bouma et al., EMBO Journal 18:5166-5174, 1999; Schwarzenbacher et al., EMBO Journal 18:6228-6239, 1999), it is known that the four CCP domains exhibit an elliptically shaped β-sandwich structure comprised of several antiparallel β-strands wrapped around a well-defined hydrophobic core containing one conserved tryptophan each. In contrast, domain V folds into a central β-spiral with two small helices and carries a distinct positive charge in the proximity of a surface-exposed loop region comprising Trp316. Domain V carries the lipid binding region within the lysine-rich sequence motif (281CKNKEKKC288) and a hydrophobic loop (313LAFW316) important to membrane binding.
From experiments designed to identify the lipid-binding site of β2GP1 (now known to be domain V), Hunt and Krilis identified an inactive form of β2GP1 that does not bind lipids and as a result did not bind lipid/β2GP1 complex-specific phospholipid antibodies (Hunt and Krilis, J. Immunol. 152:653-659, 1994; Hunt et al., Proc. Natl. Acad. Sci. (USA) 90:2141-2145, 1993). Sequence analyses of the inactive form yielded two N-terminal sequences. One of the N-terminal sequences corresponded to the N-terminus of the intact active form of β2GP1, while the other was a new sequence that started at Thr-318. The authors concluded that the inactive form was cleaved in domain V between amino acids Lys-317 and Thr-318 but was still a single polypeptide because the disulfides delineating the five domains remained intact. Although the crystal structure of the nicked protein has yet to be determined, it is reasonable to assume that the cleavage of β2GP1 at Lys 317/Thr 318 results in a dramatic change in the proteins conformation since its lipid binding properties are essentially abrogated and lipid/β2GP1 complex-specific antiphospholipid antibodies no longer bind the protein. Using the crystal structure of the intact protein, molecular modeling, and epitope mapping with monoclonal β2GP1 antibody, Matssura et al. (International Immunology 12:1183-1192, 2000) proposed that cleavage results in novel hydrophobic and electrostatic interactions in domain V that affect lipid and phospholipid antibody binding. Importantly, these changes might propagate down through the polypeptide to other domains.
Varying levels of endogenous nicked protein can be found in the plasma of certain individuals, especially leukemia patients (Itoh et al., J. Biochem. 128:1017-1024, 2000) and patients treated with streptokinase (Horbach et al., Throm. Haemostasis 81:87-95, 1999). In vitro, β2GP1 is proteolytically cleaved by enzymes that participate in the coagulation cascade; factor Xa, elastase, and plasmin all clip β2GP1 at Lys-317 and Thr-318. Moreover, the addition of urokinase to plasmin inhibitor (α2PI) depleted plasma also generated nicked β2GP1 (Ohkura et al., Blood 91:4173-4179, 1998). These observations suggest that activation of fibrinolysis in vivo induces the cleavage of intact β2GP1 by plasmin which results in the generation of the nicked protein (Blank et al., Proc. Natl. Acad. Sci. (USA) 96:5164-5168, 1999). These data taken together with the observations that β2GP1 binds endothelial cells (George et al., Circulation 99:2227-2230, 1999; George et al., Circulation 97:900-906, 1998) through annexin II (Ma et al., J. Biol. Chem. 275:15541-15548, 2000) raises the possibility that localized production of plasmin at the endothelial invasion front during wound repair could cleave β2GP1. Thus, the nicked form of β2GP1 could be generated directly on the cell surface, which would result in altered properties that directly affect endothelial cell function.
Angiogenesis, the process by which new blood vessels are formed by sprouting from pre-existing vessels, is a highly regulated process that involves endothelial cell proliferation, proteolysis of matrix molecules, self-association, elongation and migration. Angiogenic agents may be used to induce angiogenesis, or it may be the result of a natural condition. Angiogenesis is essential to a variety of normal activities, such as reproduction, development, tissue and organ growth, and wound repair, and involves a complex interplay of molecules that stimulate and inhibit the growth and migration of endothelial cells, the primary cells of the capillary blood vessels. Normally these molecules maintain the microvasculature in a quiescent state (i.e., without capillary growth) for prolonged periods which can last for several years or even decades.
While angiogenesis is essential to homeostasis (Adams et al., Genes & Development 13:295-306, 1999; Erlebacher et al., Cell 80:371-378, 1995), its timely inhibition is critical to normal wound healing and development. For example, during wound repair endothelial cells can undergo rapid proliferation, with a much shorter turnover time (Folkman and Shing, J. Biol. Chem., 267:10931-34, 1992; Folkman and Klagsbrun, Science 235:442-47, 1987). Should the balance between pro and antiangiogenic regulators go awry, uncontrolled capillary formation such as that seen in rheumatoid arthritis, diabetic retinopathy, psoriasis, retrolental fibroplasias, hemangiomas, and tumor cell growth, as well as metastasis, may result (Folkman, J., Nature Medicine 1:27-31, 1995; Ellis and Fidler, European Journal of Cancer 32A:2451-2460, 1996).
There are many diseases, categorized as “angiogenic diseases,” which are characterized by persistent unregulated angiogenesis. Unregulated angiogenesis can either be the direct cause of the particular disease, or it may exacerbate an existing pathological condition. For example, ocular neovascularization appears to be the most common cause of blindness and underlies the pathology of several eye diseases. In arthritis, newly formed capillary blood vessels can invade joints and destroy cartilage. In diabetes, new capillaries may form in the retina and invade the vitreous humor, which can cause bleeding and blindness.
The growth and metastasis of solid tumors can also be angiogenesis-dependent (Folkman, J., Cancer Res. 46:467-73, 1986; Folkman et al., “Tumor Angiogenesis,” Chapter 10, pp. 206-32, in The Molecular Basis of Cancer, Mendelsohn et al., eds. (W. B. Saunders, 1995)). Tumor cells must attract new vessels to expand locally and produce metastasis. For example, tumors that enlarge to greater than 2 mm. in diameter need to obtain their own blood supply, and can do so by inducing the growth of new capillary blood vessels. These new blood vessels become embedded in the tumor and provide nutrients and growth factors essential for tumor growth, as well as a means for tumor cells to enter the circulation and metastasize to other distant sites (Weidner et al., New Eng. J. Med., 324(1):1-8, 1991). Drugs that function as natural inhibitors of angiogenesis have been shown to prevent the growth of small tumors in tumor-bearing animals (O'Reilly et al., Cell 79:315-328, 1994). Sometimes the use of such negative regulators leads to tumor regression and dormancy even after cessation of treatment (O'Reilly et al., Cell 88:277-85, 1997). Additionally, it has also been shown that supplying inhibitors of angiogenesis to certain tumors can potentiate their response to other therapeutic regimens (e.g., chemotherapy).
During the past decade, several negative regulators of angiogenesis have been discovered. Compounds that have been reported to inhibit endothelial cell proliferation in different experimental systems include TGF-β, (Muller et al., Proc. Natl. Acad. Sci. (USA) 84:5600-5604, 1987), thrombospondin (Good et al., Proc. Natl. Acad. Sci. (USA) 87:6624-6628, 1990), IL-1 (Cozzolino et al., Proc. Natl. Acad. Sci. (USA) 87:6487-6491, 1990), IFN-γ and IFN-a (Friesel et al., J. Cell. Biol. 104:689-696, 1987), tissue inhibitor of metalloproteinase-1 (TIMP-1) (Takigawa et al., Biochem. Biophy. Res. Commun. 171:1264-1271, 1990), platelet factor 4 (PF4) (Maione et al., Science 247:77-79, 1990), protamine (Taylor and Folkman, Nature 297:307-312, 1982), fumagillin (Ingber et al., Nature 348:555-557, 1990) and angiostatin (O'Reilly et al., Cell 79:315-328, 1994).
Some of these proteins are proteolytic fragments of the same proteins that control the balance between the formation and dissolution of fibrin clots formed as a result of tissue damage and wound healing. For example, proteolysis of antithrombin III (van Boven and Lane, Seminars in Hematology 34:188-204, 1997), thrombin (Tsopanoglou et al., American Journal of Physiology 264:C1302-C1307, 1993), and plasminogen, which all play a critical role in controlling clot formation and angiogenesis, result in the production of cleaved antithrombin III (O'Reilly et al., Science 285:1926-1928, 1999; Larsson, et al., J. Biol. Chem. 276:11996-12002, 2001), prothrombin fragments 1 and 2 (Rhim et al., Biochem. Bioph. Res. Co. 252:513-516, 1998), and angiostatin (O'Reilly et al., Cell 79:315-328, 1994; O'Reilly et al., Nature Medicine 2:689-692, 1996), respectively, all of which inhibit the growth of vascular endothelial cells. Although several angiogenesis inhibitors are currently under development for use in treating diseases, there exists a need for better therapeutic options for inhibiting angiogenesis.